ContentCon: Why Structure Keeps Brands Visible in Answer Engines

The Gist

Visibility tracking. Contentstack Canoe shows brands how AI engines describe them to buyers, where those buyers get directed and what to fix next.
Markdown for machines. Time serves a stripped-down markdown version of every page to AI agents and sells labeled agent ads inside that layer.
Structure before AI. Alaska Airlines scaled from one brand to several on a single headless CMS by betting on semantic tokens and a shared content model.

CHICAGO — As brand visibility and discoverability continue to shift amid the AI boom, Conor Egan, senior vice president of product at Contentstack, has a potential solution. Last month, Contentstack launched Contentstack Canoe, an AI visibility tool that shows brands how they appear in the answers AI engines give their buyers.

(Editor's note: We had two perspectives this year at ContentCon, including Contributor Greg Kihlstrom's coverage of the AI-human content experience tradeoff and the future of digital experience platforms).

“The real meat and potatoes of it is it actually does give you tangible recommendations. It's going to constantly be monitoring [and] looking at all the responses,” Egan told CMSWire at his company's customer conference last week. “It tells you the types of pages that you might want to build. Then you can feed this into our CMS agents, and they can draft pages for you, or update existing pages to be more readable by AI.”

Egan gave a demonstration of Contentstack Canoe at ContentCon at Convene in Willis Tower last week alongside CEO Neha Sampat. A running theme throughout the event was maintaining brand visibility and discoverability in the AI era.

Contentstack Canoe aims to address this challenge directly. The tool shows brands how they appear when buyers ask AI for recommendations, where buyers are directed and gives recommendations on how brands can improve their AI visibility.

How Contentstack Canoe Maps Buyer Questions by Stage

The tool scans the site to identify the business, products, buyer personas and up to 10 competitors. Users can review that information before Contentstack Canoe generates the questions their buyers are likely to ask. Questions are also tagged by buying stage — from awareness through decision — and users can edit them.

“We can feed engines just content without all of that decoration,” Egan said. “AI likes to get definitive sources, [it] likes to have small encapsulated answers that it can pull. It doesn't want to read through a super complex HTML site with a bunch of animations and everything.”

Add CMSWire as a Preferred News Source

Contentstack's Recent Platform Releases

The launch of Contentstack Canoe follows the September release of a set of updates to its Assets add-on that give content teams more control over metadata, automate tagging with AI and surface duplicate images for cleanup. The company also introduced self-service account creation and subscription management, opening the platform to smaller teams and individual builders.

Last year, Contentstack launched Agent OS, an AI-powered headless CMS that delivers real-time personalization, workflow automation and context-driven customer experiences.

What Does Contentstack Canoe Do for Brand Visibility?

Contentstack Canoe scans a brand's site, generates the questions buyers likely ask AI engines, tags them by buying stage and returns recommendations that Contentstack's CMS agents can turn into new or updated pages.

Time and Alaska Airlines Build Content for Machine Readers

Egan said that being able to publish markdown versions of a site on Contentstack Canoe is optimized for discoverability, which other brands at ContentCon reflected in their success stories.

Time Serves Markdown Pages to AI Agents

For example, Michael Mraz, GM and chief product officer at Time Magazine, talked about how the company creates markdown versions of all of its pages, so that it's much easier for AI to process.

“Time sees more bot traffic than human traffic, so we're building for that audience now too. We now serve a markdown version of every page we publish, exclusively for agents, a stripped-down format designed to be read by machines,” Mraz said in his keynote.

The 102-year-old magazine partnered with AI ad tech startup Mobian to “create an entirely new form of ad inventory: agent ads. These are verified brand facts, clearly labeled, embedded directly into that markdown layer, and measured by visibility, favorability and accuracy in AI search.”

Time drew media attention in August after serving Ads in its markdown version of web pages. Digiday reported one LLM, Perplexity, blocked what it called a deceptive practice. Time disagreed its practices are deceptive in a follow-up interview with The Media Copilot on LinkedIn, calling it a transparent way to present the information to an AI crawler.

Alaska Airlines Runs Multiple Brands on One CMS

In terms of brand individuality, speakers from Alaska Airlines explained how they expanded their digital operation from a single airline brand to three brands in seven languages while maintaining brand characteristics.

This journey started in 2022, when they were one brand and the site was two languages. They needed to find a way to effectively add another brand without having to duplicate the website again. This led them to use Contentstack.

“The question landed on our team: Can our new platform carry two airlines, two brands, one promise? We chose to preserve the soul of each, not to blend them away,” said Lindsey O’Sullivan, manager of digital guest experience at Alaska Airlines. “Here's the part that still surprises people: it's the same CMS, it's the same page types, it's the same component library, and it's the exact same team sitting in here today as before we started.”

One CMS now feeds their website, Alaska Airlines App and in-flight entertainment.

Senior product manager Brandon Shields said the team made three “bets” when making the transition: a semantic design token layer to their design system; using a headless API-first CMS as one source of content with every channel pulling from it; and decoupled UI components that live outside of the CMS so they can evolve independently.

Those bets paid off. This model was successful, and brands kept requesting to join. Earlier this year, the airline added new routes to Rome and London.

“What's different is the brand layer and the content," Shields said. "But once we said yes to one new brand, the requests kept coming.”

Start With Semantic Design Tokens

They advised attendees to start with a semantic tokens layer and to have all of the brands use the same model instead of having brand-specific content models.

“Our designers can now use content models that are live in production and our token layer to build new experiences on the web, in the app, or in places we haven't even imagined yet. We didn't plan for that, but the foundation made it possible. Structure first, AI second,” Shields said.

How Are Time and Alaska Airlines Structuring Content for Machines?

Time serves a markdown version of every page for agents and embeds labeled agent ads in it, a practice Perplexity blocked as deceptive. Alaska Airlines feeds its website, app and in-flight entertainment from one headless CMS built on semantic tokens and a shared content model.

Why AEO Tools Still Lag Behind AI Discovery

In terms of challenges for marketers trying to produce content that gets found in LLMs, Egan said that stems partially from the newer nature of AEOs as opposed to SEOs.

“We're just in this awkward middle part where [AI] is very important, and there's very few tools for marketers and people that need to be navigating it.” He emphasized that while discovery has moved to AI, decisions are still with humans.

Where Answer Engine Tools Go Next

In the future, Egan said he foresees that big search engines are going to provide more tools and standards to work with.

“We're just in a very early period, and I think what will happen is as these engines give more tools to marketers and technologies like us, there will maybe be less of the AI having to make stuff up and closing gaps because they all want their agents to give accurate information,” he said.

In terms of what brands will make it past the awkward phase, he said those that embrace the shift from SEO to AEO are more likely to find longevity than those that don’t.

“Every company needs to be ready to change how they work, and if they're resistant, they're less likely to get through it,” he said. “This is really an opportunity for us to take our strengths and make them relevant for AI.”